One third of the 500,000 individuals who suffer a stroke each year have mobility impairment. The overall goal of this research is to construct a scientific foundation for understanding bilateral coordination deficits in persons with hemiparesis that is based on neuromotor and biomechanical control principles and will be applicable to gait rehabilitation. Elucidating deficits in the bilateral coordination of rhythmic locomotor behavior is a necessary step towards developing sound, scientific principles for choosing effective strategies for the rehabilitation of hemiparetic stroke patients. Since hemiparetic locomotion is asymmetrical, the two legs experience substantially different mechanical loading and sensorimotor activity. Experiments on neurologically healthy individuals have shown that the locomotor pattern expressed by each leg depends not only on the mechanical load, but also on the sensorimotor activity of the contralateral leg independent of the load. Coordination deficits in the paretic leg may, therefore, result in part from the sensorimotor activity of the non-paretic leg. The proposed study will employ a novel pedaling apparatus to decouple the effects of mechanical load and contralateral sensorimotor activity in ways that cannot be achieved in gait experiments. Each leg is connected to its own crank, which is connected to a servomotor that can be programmed to apply a load dependent on, or independent of, the kinematic and kinetic state of the contralateral leg. Computer simulations and experiments performed on individuals with post-stroke hemiparesis and age-matched neurologically healthy subjects using this pedaling apparatus will test these hypotheses: 1) The motor pattern of the non-paretic leg differs from he neurologically healthy pattern primarily to compensate for the lateral mechanical load experienced by the non-paretic leg; 2) The motor pattern of the paretic leg differs from the neurologically healthy pattern in ways that are not related to the mechanical load experienced by the paretic leg; 3) The motor pattern of the paretic leg is inappropriately influenced by the sensorimotor state of the contralateral leg (e.g., muscular activity and afference related to loading and movement.)